Dell has worked with more than 20 quantum companies over the past decade, and HPE used its Discover 2026 show to announce partnerships with Intel, IQM Quantum Computers, Quantinuum, QuEra, and Rigetti. The two system makers have reached the same verdict on quantum computing: it enters the datacenter as an accelerator, not as a revolution. That verdict now shapes their product roadmaps.

How HPE frames the integration job

HPE is building a platform to integrate quantum systems of multiple modalities with its Cray supercomputers, including testbeds for hybrid algorithm design, software interoperability, and benchmarking of HPC-quantum workflows. Trish Damkroger, senior vice president and general manager of HPC and AI infrastructure solutions at HPE, put the company's role plainly.

"We will not make the quantum devices, just like we don't make the silicon," Damkroger told The Next Platform. "I look at quantum as another accelerator. We will be putting the network to it. You've got to bring that all together."

Her reference point is historical. "We're not at the Cray-1 level yet, but we're getting there," she said, meaning the quantum market sits where supercomputing sat when the Cray-1 launched in 1975. Oak Ridge National Laboratory, home of the Cray-based Frontier supercomputer, can already order quantum capacity over the cloud. Damkroger expects labs to move to owned hardware.

Where on-premises systems already run

Oak Ridge moved first. In June the lab announced that Pathfinder, a 20-qubit IQM Radiance system, went operational. It is the first quantum system the national lab has bought and the first IQM has sold in the United States.

IQM, based in Finland, has sold 23 systems worldwide and focuses on on-premises datacenters. Its Halocene family, aimed at error-correction research, launched in November 2025 with a 150-qubit system. CSC, the IT Center for Science in Finland, will deploy the Halocene H4 next year as part of the LUMI AI Factory. D-Wave, historically a cloud operator through Leap, began selling its Advantage systems directly last year.

Ron Bewtra, IQM's country director in North America, wrote recently: "My view, shaped by what I've seen in the U.S. market, is that the most credible near-term path runs through HPC."

What Dell showed on the floor

Dell's position mirrors HPE's. "There are no quantum computers that are only built with QPUs," said John Roese, global CTO and chief AI officer at Dell. "They are all hybrid quantum systems. Our strategy is to be their best friend, to treat it like an accelerator, just like GPUs."

At Dell Technologies World 2026, the company displayed a prototype of Equal1's RacQ quantum computer, the latest generation of the Bell-1 Server. The system is a concrete example of where quantum hardware is heading:

• Silicon-spin qubits fabricated on standard CMOS, with all system components in a single silicon package
• Roughly 1.6 kilowatts of power draw and 400 kilograms in weight, fitting a standard Dell 42U frame
• An integrated cryocooler holding 0.3 Kelvin, removing the need for external cryogenics
• Rack-mountable in a standard 19-inch datacenter rack

The demonstration ran classical workloads on Dell PowerEdge servers, with the quantum machine handling the mathematically heavy subroutines.

Scaling RacQ from five qubits toward 1,000 or a million requires two engineering feats, Roese said. The die where the qubits live must grow from about 3 millimeters squared to 11 millimeters squared inside the 300-millikelvin environment, and the control system must shrink from a set of boxes into a processor using Cryo-CMOS, CMOS compute running inside a super-cooled environment. "That's a whole different technology," he noted.

Why co-location wins

The commercial logic of hybrid systems comes down to cost. Burns Healy, Dell's quantum infrastructure lead, said at a roundtable: "If you want real near-term quantum advantage, you're not going to have that not be co-located with HPC because, relative to qubits, the silicon is the cheap part."

Healy also argued that predictive machine learning will help organizations decide when to switch on a GPU rather than a QPU, CPU, or FPGA. Different quantum processor technologies will remain better at different workloads, which complicates scheduling.

Roese framed the end state in familiar terms: roll an Equal1 system next to an Nvidia NVL72 rack, connect the two, and you have the basic compute for a CPU-GPU-QPU system. The sector's near-term future, on this evidence, is quantum hardware sold as a rack component, benchmarked and serviced by the same OEMs that already own the classical datacenter.